Op-Ed: Our View on Value Frameworks in Oncology: Proposing Principles for Value Framework Development
By Lee Schwartzberg, MD, David Ettinger, MD, Mohammad Jahanzeb, MD, Gregory Otterson, MD and David Waterhouse, MD
In the past year, Value Frameworks have been introduced as potential tools for policy-setting and decision-making in oncology by organizations including ASCO, NCCN, and ICER. We recently reviewed a report issued by ICER, the Institute for Clinical and Economic Review, regarding Non-Small Cell Lung Cancer (NSCLC). This report is the basis for a meeting and voting by ICER’s committee that will take place on Thursday, October 20, 2016, in St. Louis, MO.
For us as practicing oncologists and lung cancer researchers, this report has raised serious concerns regarding ICER’s ability to interpret clinical evidence and reach conclusions on drug value that are scientific, comprehensive, and unbiased.
Based upon our reviews of this report, ICER appears to represent a perspective that is less oriented towards patient benefit than towards motivations that would limit patient access to new therapeutic options. ICER’s clinico-economic methods include approaches and metrics that, due to their singular focus on population-level health, would likely fail patients on an individual, clinical needs basis. ICER’s philosophy appears to be similar to that of NICE in the UK, whose limitations placed on drug access have been correlated with lower cancer survival rates in the UK compared to the rest of Western Europe.
In response, we propose the following principles as the start of a discussion on standards and best practices for Value Framework development, particularly as these frameworks have the potential to impact clinical practice and the patients we serve.
Principle 1: Disease Experts as Evaluators and Authors
In any Value Framework, clinical benefits and toxicities should be assessed by clinicians who have extensive and active expertise, in research and in practice, of the disease area and treatment options being evaluated. For example, a report on lung cancer should have practicing physicians who are treating patients with the drugs in question as primary evaluators of such drug technologies and authors of any publications.
Principle 2: Patient-Centered Endpoints, Conclusions, and Definitions of Value
Patients must come first. The methods, processes, value definitions, and conclusions of any Value Framework should be patient-oriented so that the policies and decisions it informs serve patients’ best interests. Likewise, primary evaluative endpoints such as Overall Survival (OS) or Progression Free Survival (PFS) for any analysis should be scientifically and clinically selected to reflect the clinical benefit of primary importance to patients. The selection of trials, datasets, and studies for inclusion in any analysis should permit a full assessment of outcomes achieved and not exclude valid studies that could contribute to accurate evaluations. Metrics selected to define value should also be validated to ensure they reflect current and relevant patient definitions of value.
Principle 3: Rigorous Methodologies Reflecting Evidence Based Medicine
Evaluation approaches that use cross-trial comparisons, comparisons of clinical outcomes or health utility states across different populations, or extrapolations from one indication to another without an evidentiary basis are methodologically perilous and fraught with the potential for wrong conclusions. Therefore, cross-trial methodologies that lead to comparing or aggregating dissimilar patient populations should be avoided; such comparisons are likely to produce confounding results, particularly in complex areas involving biomarker selection, biomarker measurement, histology-specific populations, or differing mechanisms of action. Methodologies that do not reflect rigorous, scientific and statistical best practices for evidence based medicine or true meta-analysis severely limit the validity and applicability of conclusions derived from such methodologies.
Principle 4: Continuous Review and Revision
Value Frameworks should be viewed and developed as continuous works in progress, reflecting an ever-changing treatment landscape. Criteria and processes for timely and prompt review of new scientific data, as well as revision and publication of new conclusions, should be integral components of any Value Framework effort.
Principle 5: Peer Review and Authorship to Scientific Standards
Organizations authoring Value Frameworks should submit their methodologies and assessments for publication in relevant peer-reviewed journals. Publications or reports should adhere to authorship standards for scientific documents and avoid general characterizations and terms. Terminology, especially in explaining results or outcomes, should be proximately and directly explained in a quantitative fashion and referenced by study results. Assumptions and methods for any assessment should be transparently documented, explained, referenced, and evaluated for potential bias.
We support these principles and wish to see them widely communicated and adopted as potential best practices in Value Framework development. We also invite our colleagues, our patients, and other stakeholders in cancer care delivery to communicate their perspectives on critical principles for Value Frameworks as we move this field forward.